Test marketing (NPD Stage 5)
Category: Marketing
The manner in which new products are tested in the real competitive marketplace takes many forms from some limited market test experiments («limited testing») to «full test marketing».
The decision as to whether to test market at all and if so whether by «limited» or «full» testing, is determined by seven main considerations:
— Levels of confidence in the product concept.
— Indications already received about possible market reaction.
— Whether the product is durable.
— Competitive circumstances.
— The significance of the new product in the overall portfolio.
— The acceptability of the testing procedures.
— Costs of the tests.
There may be grounds for feeling fairly confident about projections that had been made about the product’s prospects. The company may have extensive, sophisticated knowledge and experience of operating in the market segment. If that was backed by preliminary market research, and if there were strong favourable indications from concept and product testing then subsequent testing may be foregone. But even with this set of circumstances there could be an argument for some types of testing especially connected with the marketing plan. For example, there could be great uncertainty surrounding the extent to which various advertising approaches correctly underscore the positioning strategy or conflicting views about alternative sales promotion plans.
In some durable products/markets many kinds of experimentation may not be at all feasible because of the cost and lead-time in manufacturing. Full-scale tooling cannot proceed simultaneously for several possible new washing machines or cars, in order to mount a test market. Apart from anything else the company’s reputation could be damaged if it became known that it was putting experimental products on the market. Such firms are forced to rely upon earlier testing.
Guesses about competitive reaction can also affect test decisions. It may be known that several companies are rushing to bring out similar new products. Market experiments would delay national launch and give competitors the chance to evaluate rival products and plans It would be expected that competitors would monitor any experiments, commissioning special market research on their rival’s new product. Disruptive spoiling tactics may also be used, with competitors buying up product, putting in extra promotional activity in the test area dropping prices or attempting to influence the trade against the newcomer. Aggressive competitive reaction could preclude or curtail test marketing and make more limited experiments attractive.
The new product might be of variable significance in the overall product mix or portfolio. If a major investment were at stake, there would be a greater reason to employ a wide range of full market tests. If the product was only a variation of an existing product item and much smaller investment was being risked then less testing would be likely.
Testing also imposes costs. With test marketing unit production costs would be high and management costs might also be of a high order because of the attention give to high profile endeavours such as new product introductions.
For these reasons attempts to improve testing procedures and to allow much earlier predictions are always in hand. Reducing the risk in new product launches, within acceptable time and cost limits, is the goal.
One form of limited testing involves conducting tests in a small number of retail outlets that agree to participate in a test. Panels of consumers are recruited at these locations and they record their purchases in a weekly diary. Electronic data capture has allowed sophisticated extensions to this basic idea. A panel of regular shoppers at a store is recruited and given plastic cards, which are scanned at the checkout each time they visit. Demographic and attitudinal data are obtained from those cooperating. Test products can be introduced to the store and sales to panel members monitored. Special analyses of penetration, repeat and buying volume rates within demographic or attitudinal grouping are therefore possible. There could also be some degree of control over the promotion of test items possibly using direct mail to panel members or special editions of free newspapers. In the USA cable television is used to screen test advertisements to such shopping panels and their viewing recorded electronically. This provides integrated product and media data. For comprehensive coverage see Crawford.
Although there has been considerable development of simulated or limited test marketing procedures the full scale test marketing exercise is still widely employed. This is because full test marketing can do things other tests cannot do. For example the very act of placing the new product on sale in a reasonably large area provides a much more natural «in vivo» test than does the rather «in vitro» limited test. The preoccupation in limited testing is the reactions of consumers. This misses a vital set of reactions which will be just as powerful in determining the product’s prospects: those of trade. There might also be significant problems in the logistics of bringing out a new product which need to be evaluated in a natural setting. Additionally, the impact of the full marketing mix can only be realistically evaluated in a fairly large test operation. Fuller information on the key variables to be used in estimating sales levels may also be collected. All this argues for test marketing despite its cost, the time it takes and the exposure of company plans.
Davis has put forward a threefold classification of the purposes in test marketing (table below). The pilot launch checks the logistics, marketing mix testing investigates the returns to alternative marketing efforts and sales testing yields insight into possible national sales levels. All three could be of equivalent moment but frequently there would be more uncertainty about one of these areas and so most attention would be paid to that.